Among nuclear receptors, PPAR (Peroxisome Proliferator Activated Receptor) is known to have three subtypes, which are PPARα, PPARγ and PPARδ (Nature, 1990, 347, p 645-650, Proc. Natl. Acad. Sci. USA 1994, 91, p 7335-7359). PPARα, PPARγ and PPARδ have tissue-specific functions in vivo and different regions for expression. PPARα is mainly expressed in the heart, kidney, skeletal muscle and large intestines in humans (Mol. Pharmacol. 1998, 53, p 14-22, Toxicol. Lett. 1999, 110, p 119-127, J. Biol. Chem. 1998, 273, p 16710-16714), and is involved in β-oxidation of peroxisome and mitochondria (Biol. Cell. 1993, 77, p 67-76, J. Biol. Chem. 1997, 272, p 27307-27312). PPARγ is expressed in the skeletal muscle at a low level, but mainly expressed in the adipose tissue to induce the adipocyte differentiation and to store energy as a form of fat, and is involved in homeostatic regulation of insulin and glucose (Moll. Cell. 1999, 4, p 585-594, p 597-609, p 611-617). PPARδ is preserved evolutionarily in mammals including humans and vertebrates including rodents and sea squirts. The previous studies confirmed that PPARδ plays an important role in the reproductive cell expression (Genes Dev. 1999, 13, p 1561-1574) and has physiological functions of differentiating neuronal cells (J. Chem. Neuroanat 2000, 19, p 225-232) in central nervous system (CNS) and wound healing with anti-inflammatory effect (Genes Dev. 2001, 15, p 3263-3277, Proc. Natl. Acad. Sci. USA 2003, 100, p 6295-6296). Recent studies also confirmed that PPARδ is involved in the adipocyte differentiation and lipid metabolism (Proc. Natl. Acad. Sci. USA 2002, 99, p 303-308, Mol. Cell. Biol. 2000, 20, p 5119-5128). For example, PPARδ activates the expression of key gene involved in β-oxidation in fatty acid catabolism and uncoupling proteins (UCPs), the gene involved in energy metabolism, which brings the effect of improving obesity (Nature 2000, 406, p 415-418, Cell 2003, 113, p 159-170). The activation of PPARδ increases the HDL (High Density Lipoprotein) level, improves type 2 diabetes without weight changes (Proc. Natl. Acad. Sci. USA 2001, 98, p 5306-5311, 2003, 100, p 15924-15929, 2006, 103, p 3444-3449), and favors the treatment of arteriosclerosis by inhibiting the gene associated with arteriosclerosis (Science, 2003, 302, p 453-457). Therefore, PPARδ ligand can be developed as a drug for the treatment of metabolic diseases such as obesity, diabetes, hyperlipidemia and arteriosclerosis.
PPARδ regulates mitochondria biosynthesis. When PPARδ was artificially over-expressed in the mouse muscles, mitochondria biosynthesis was increased and Type I muscle fiber was increased significantly, in addition to the increase of fatty acid β oxidase. Therefore, constant running time and distance were respectively 67% and 92% increased, compared with wild type mouse (PLoS Biology, 2004, 2:e294). The increase of mitochondria biosynthesis has a positive effect on the enhancement of brain functions. If mitochondria in brain cell is destroyed by oxidative stress, memory decreases significantly (Proc. Natl. Acad. Sci. USA 2002, 99, p 2356-2361). Dementia, Alzheimer's and Parkinson's disease are the representative degenerative diseases, which demonstrate significant decrease of learning and memory. Therefore, the mitochondria proliferating agent developed in the present invention not only contributes to the improvement of memory, but also can be developed as a therapeutic agent for Alzheimer's and and Parkinson's disease.
Synthetic PPARδ ligands developed, so far, have less selectivity, compared with other PPARα and PPARγ ligands. The early selective ligand was L-631033, developed by Merk (J. Steroid Biochem. Mol. Biol. 1997, 63, p 1-8), which was produced by introducing a functional group being able to fix side chain based on its natural fatty acid morphology. The same research team reported later more effective ligand L-165041 (J. Med. Chem. 1996, 39, p 2629-2654), in which the compound known as a leukotriene agonist is functioning to activate human PPARδ. This compound exhibited great selectivity to hPPARδ, which is 10 times the selectivity to PPARα or PPARγ. And EC50 of the compound was 530 nM. Other ligands L-796449 and L-783483 have improved affinity (EC50=7.9 nM), but barely have selectivity to other hPPAR subtypes.
Glaxo-Smith-Kline reported GW2433 (Chem. Biol. 1997, 4, p 909-918), the PPARα activator, which is Y-type ligand having a similar structure to the crystal structure of the PPARδ ligand pocket. Unlike the conventional ligands known so far, this ligand has Y-type structure containing benzene ring, which favors spatial binding to the PPARδ ligand pocket. However, this ligand is a double-active ligand having activity to hPPARα as well, suggesting that selectivity to PPARδ is reduced. The PPARδ selective ligand GW501516 ([2-methyl-4-[[[4-methyl-2-[4-(trifluoromethyl)phenyl]-1,3-thiazol-5-yl]methyl]sulfanyl]phenoxy]acetic acid), recently developed by GlaxoSmithKline, exhibits much better physiological effect than any other ligands previously developed (Proc. Natl. Acad. Sci. USA 2001, 98, p 5306-5311).

The GW501516 has excellent affinity (1-10 nM) to PPARδ, and excellent selectivity to PPARα or PPARγ as well, which is at least 1000 times the selectivity of earlier ligands.
However, the PPARδ activity induced by all the ligands developed so far is only resulted from 30-40% of total ligand-binding pockets.
WO 2001-00603 applied by Glaxo group describes the compound represented by the following formula A containing the GW501516 as a selective activator of PPARδ. However, this description includes only a part of the test results of GW501516 using Rhesus model.

[Wherein, R′ is CF3 or F, R″ is H, CH3 or Cl, and R′″ is H, CH3 or CH2CH3.]
The thiazole compound represented by formula B as a PPARδ selective activator has been described in WO 2002-62774 applied by Glaxo group.

WO 2003-072100 applied by Eli Lilly describes a pharmaceutical composition for selective regulation of PPARδ containing the compound represented by the following formula C.

However, the description only declares that the composition has been prepared, which is produced as a racemate not as an optical isomer comprising two types. And the document only describes M++1 value of mass spectrometry of the produced racemate, which is confirmed by 1H-NMR and acts as a selective activator of PPARδ, but does not mention about the pharmacological effect as a selective activator of PPARδ.